An evaluation device includes: a memory; and a processor coupled to the memory and the processor configured to execute a process, the process comprising: generating a plurality of sets from a first population of biometric data collected from a plurality of users, a number of biometric data of the plurality of sets being smaller than that of the first population, at least a part of the plurality of sets being different from each other; estimating a probability density distribution of similarity degrees of different people's pairs of each of the plurality of sets; calculating a false acceptance rate with respect to each similarity degree threshold on a basis of the probability density distribution, with respect to each of the plurality of sets; and estimating a combined false acceptance rate from a statistic amount of false acceptance rates of the plurality of sets calculated by the calculating.